Add tests

[jranke]

Currently there are no tests driven from tests/ either directly or via a test runner in the package but only tests relying on help page examples that do not check the output of the compiled functions.

[jranke]

Thanks, that's more precise now.

[eddelbuettel]

The advantage of tinytest is that it is very easy (and always possible) the run the tests of an installed or source package. I actually do that a lot (often via tt.r which is a convenience wrapper for the shell) --- but I also appreciate the fact that the package itself is zero-depends. A trade-off.

[jranke]

You decide, if you like tinytest I will give a go.

[eddelbuettel]

It's worth a try. It is purely additive: add a Suggests:, copy the caller into tests/ and add a file. See how it goes. Be incremental.

[jranke]

BTW, the test fails in the same way when run in the CRAN version 0.3.16.

It's enough to put the following code in the tests/ directory:

library(inline)

n = 10L
x = 1:10

## A simple Fortran example - n and x: assumed-size vector
code <- "
      integer i
      do 1 i=1, n(1)
    1 x(i) = x(i)**3
"

cubefn <- cfunction(signature(n = "integer", x = "numeric"), code,
   convention = ".Fortran")

and I get

* checking tests ...
  Running ‘cfunction.R’
 ERROR
Running the tests in ‘tests/cfunction.R’ failed.
Last 13 lines of output:
    4:       DOUBLE PRECISION x(*)
    5: 
    6:       integer i
    7:       do 1 i=1, n(1)
    8:     1 x(i) = x(i)**3
    9: 
   10:       RETURN
   11:       END
   12: 
  
  Compilation ERROR, function(s)/method(s) not created!
  Error in compileCode(f, code, language, verbose) : 
    Error in file(filename, "r", encoding = encoding) :   cannot open the connectionCalls: local ... eval.parent -> eval -> eval -> eval -> eval -> source -> fileIn addition: Warning message:In file(filename, "r", encoding = encoding) :  cannot open file 'startup.Rs': No such file or directoryExecution halted
  Calls: cfunction -> compileCode
  Execution halted
* checking PDF version of manual ... OK
* DONE

Status: 1 ERROR

As mentioned above, this is with the released version.

[eddelbuettel]

🤷‍♂️

I really think it's a known (to calls of R CMD check) issue I have seen before / we all have seen before but I don't have time right now, unfortunately, to google or research this for you. Of course I could also be flat out wrong. Who knows?

Back to real work, and end to the staccato over this topic. Later...

[eddelbuettel]

Thanks for the PR. I should have done that a while ago but because some low-dependency packages (such as Rcpp) also have a Suggests: inline` I was overly hesitant.

I also just increased the version number to signal this 'in-between releases' status and made a fairly routine update to the CI script I use (and similarly update in a few other places). It would be great if you could rebase your other branch to the main one here.

Edit: And I was of course too optimistic as switching to a matrix build with macOS promptly goes belly-up over Fortran. I'll disable the macOS part.

[jranke]

You're welcome. My pleasure!

I don't have any macOS so I can't do anything there.

I rebased my other branch as suggested and will add tests for the new stuff there... Having tests will really make it easier to see if I goof up while trying to fix things! The tests already relevaled two things I overlooked (one was in SetCMethod which I hadn't looked at). I currently do not plan to add tests for cxxfunction. I saw it uses compileCode() from R/cfunction.R but I think I don't need to touch that.

[eddelbuettel]

No macOS here either, and that is the part that Travis CI will make us pay so off it is. [ I am currently reworking / rebranding my Travis setup script as it works unchanged at Travis, GitHub Actions, Azure Pipelines, ... so we could always run it somewhere else. ]

And yes, having tests is good. I'll add something simply for cxxfunction 'in due course'. No rush. The "mutex" is yours, make your changes; and I appreciate that you are being careful and diligent about it. And in the process don't forget to add yourself in DESCRIPTION as an author.